A circuit forming an active rc filter for use as a band-stop filter in high and very-high frequency domains, having a second order transfer function, comprising a circuit that includes a building block for performing a filtering and amplifying function, and a building block S3 for performing the summation VS of the circuit output signal and the signal VE applied to the input E of the circuit, characterized in that the filter building block is formed by a series arrangement of two filter building blocks F1 and F2 each having a first order all-pass function, and in that the circuit also includes an amplifier building block A1 in a series arrangement with the filter building blocks F1, F2, inserted between the output of the latter building blocks and the output of the circuit, in that the input signal VE is applied on the other hand to the summation building block S3 through an amplifier A2, and in that the summation building block S3 is formed by coupling the outputs of the amplifier building blocks A 1 and A2 for forming the circuit output S at which the output signal VS is available and in that each building block F1 and F2 comprises a phase-shifting bridge associated to an amplifier stage.
|
1. An active rc filter circuit, comprising:
a first filter receptive of the filter circuit input signal and having a first order all-pass filter function; a second filter having a first order all-pass filter function connected in series with said first filter; a first amplifier connected for receiving the output of said second filter; a second amplifier connected for receiving the input of the active filter circuit; and summing means for summing the respective outputs of said first and second amplifiers for developing the output of the active filter circuit.
2. An active rc filter circuit according to
3. An active rc filter circuit according to
4. An active rc filter circuit according to
5. An active rc filter circuit according to
a phase-shifting bridge circuit, said phase-shifting bridge circuit comprising a pair of parallel branches between two nodes defining the input nodes of said bridge circuit, a first of said branches comprising a first resistor and a first capacitor in series between said bridge input nodes, the second of said branches comprising a second capacitor and a second resistor in series between said bridge input nodes and in an order opposite that of said first capacitor and said first resistor, the respective capacitor and resistor comprising said first branch and said second branch joining at a node, and the nodes of each branch together defining the output of said phase-shifting bridge circuit; and an amplifier for amplifying the output signal from said phase-shifting bridge.
6. An active rc filter circuit according to
7. An active rc filter circuit according to
8. An active rc filter circuit according to
|
|||||||||||||||||||||||||||||||
The invention relates to a circuit forming an active RC filter to be used as a band-stop filter in the high and very-high frequency domains, having a second order transfer function, comprising a circuit that includes a building block for performing a filtering and an amplifying function, and a building block for performing the summation VS of the circuit output signal and the signal VE applied to the input of the circuit.
The invention is applied to the realisation of integrated circuits forming filters, in the high and very-high frequency domains, that can be used, for example in one embodiment, as band-stop filters in frequency doublers for rejecting the unwanted signal at the fundamental frequency or, in another embodiment, as a band-stop filter having differential inputs and outputs; or, in yet another embodiment, be particularly suitable for a fine adjustment of the rejection.
An active all-pass filter is already known from the prior art from the publication by J. TOW, in "IEEE Spectrum, December 1969, pp. 103-107", entitled "A Step-by-Step Active-filter Design".
This publication describes, among other things, an all-pass filter realisation having a second order transfer function, which filter is shown in FIG. 8 of the said document. This circuit comprises three operational amplifiers arranged in series and is looped back. A fourth operational amplifier performs the summation of the output signal of the first operational amplifier and the input signal of the filter circuit
The operational amplifiers that are used for constituting this circuit have a gain which is most certainly infinite at low frequencies, but which becomes very low at high and very-high frequencies. Therefore, the circuit known from the said document has the disadvantage of not being suitable for use in these frequency domains.
Furthermore, the looped back circuit has the disadvantage to cause the circuit to oscillate in certain conditions.
FIG. 6 of the said document also shows an active band-pass filter that has similar characteristic features and thus exactly the same disadvantages.
The circuits known from the said document further have the following disadvantages: on the one hand, they are formed by a large number of transistors, which:
causes manufacturing to be costly in many cases
requires a large surface and is disadvantageous for use in integrated circuits,
entails high consumption and is also disadvantageous for the use as mentioned above.
On the other hand, their characteristic frequency is not adjustable.
Finally, the capacitors used in this circuit have considerable dimensions and render this circuit not integrable
Therefore, the present invention has for its object to provide an active filter circuit that allows to get rid of these disadvantages and that specifically:
can operate at high or very-high frequencies;
and in the case of the use at very-high frequencies is easy to integrate, requires a small surface and has little consumption;
exhibits characteristic features so that the rejection frequency and the rejection are adjustable and easy to control;
can admit a differential input signal and a differential output signal.
According to the invention, this object is achieved by means of a circuit as described in the preamble, characterized in that the filter building block is formed by a series arrangement of two filter building blocks F1 and F2 each having a first order all-pass filter function, and in that the circuit also includes an amplifier building block A1 in a series arrangement with the filter building blocks F1, F2 inserted between the output of the latter building blocks and the output of the circuit, in that the input signal VE is applied on the other hand to the summation building block S3 through an amplifier A2, and in that the summation building block S3 is formed by coupling the outputs of the amplifier building blocks A1 and A2 for forming the circuit output S at which the output signal VS is available
The invention will be better understood by means of the following description illustrated by the annexed drawings, in which:
FIG. 1 represents the building block diagram of the circuit according to the invention;
FIG. 2a represents an embodiment of the circuit according to the invention by means of field effect transistors;
FIGS. 2b and 2c, respectively, represent the curves of the relationship between the output and input voltages of the circuit versus the frequency, and of the phase shift φ+π of the output signal and the input signal versus the frequency;
FIG. 3 represents a variant of the circuit according to the invention;;
FIG. 3b represents another variant of the circuit according to the invention.
In order to clarity the description of the invention it should first be recalled, that the transfer function of a second order all-pass active filter is (p2 +1)/(p2 +aBp+1): see the Table on paqes 94-95 of the publication entitled "Filtres Actifs" by Paul BILDSTEIN in the "Editionsde la Radio" (9 Rue Jacob, PARIS, FRANCE).
For an ideal second order band-stop filter this transfer function can be described according to the relationship (1): of the Table I, in which relationship:
s is the Laplace variable that is derived from the variable p used in the aforementioned publication (p being the lower Laplace variable) by the relationship (2) of the Table I, also utilizing relationship (3) of Table I.
The other parameters of the transfer function (1) are defined in the abovementioned publication where ω1 is the rejection frequency for which the function (1) is absolutely F(jω1)=0 and thus for which the rejection is complete;
Q0 is the quality factor or selectivity actor which indicates the steepness of the slope of the curve |F(jω)|in the neighbourhood of ω1 ;
ω0 is the angular frequency associated to the poles of the transfer function. Here it has been chosen to make the drive frequency ω1 correspond to the drive angular frequency ωp of the above-referenced publication, with in this publication ω1 =ω0.
G is the gain of the low-frequency circuit (if ω→0 and S→0, the factor G is reduced to 1, and F(jω)→G).
The transfer function of a non-ideal circuit, but having a real fixed structure, for which no requirements are made on the value of the elements, in the most simple case is of the second order and is expressed by the relationship (4) of Table I.
For practical reasons of simplicity one is here interested in the case in which ω1 =ω0 whatever the values of the filter elements are, and in which,
Q1 is the rejection factor,
|Q1 |>Q0 so that the filter operates in the band-stop mode, which leads us to two cases:
Q1 >Q0
and
Q1 <=Q0.
Thus, there appears a novel filter parameter which is the rejection R of the filter provided by (he relationship (5) of the Table I, if assuming that ω1 ≃ω0.
The greater the absolute value of Q1 (denoted Q1), the better is the rejection R. At the limit value the rejection R is complete when Q1 is infinite.
Well, the real band-stop filters can be subdivided into two categories:
the first category corresponds to the case in which Q1 is finite, and greater than 0 (Q1 >0) whatever the filter elements. In these conditions, the filters cannot realise a total rejection. The maximum value Q1 is the one that determines the filter limits. Actually, for a given value Q1, if the value Q0 enhances, the rejection R is diminished.
the second category corresponds to the case in which the fact that Q1 is greater or smaller than 0 contrarily depends on the filter elements. In this case there are particular element values such as (1/Q1)=0, where the rejection R is complete. The transfer function G(s) provided by the relationship (4) is thus reduced to the function F(s) provided by the relationship (1).
The filters belonging to the second category especially have the property of being realised by choosing the elements such that Q1 =-Q0 (and always ω1 =ω0), so that the function (4):
|G(jω)|=G is constant.
Thus, the filter becomes of the second order all-pass type. The phase difference between the output signals and input signals passes from 0 to -2π, when ω varies from 0 to infinity.
The main object of the invention is to realize active filters comprising the least possible number of elements in order to promote integration, and operating at high and very-high frequencies.
To this end, the integrated circuits of the filters according to the invention are advantageously realized with gallium arsenide (GaAs).
Furthermore, according to the invention, one has chosen to realise active band-stop filters of the second category, which permits to attain the following advantages:
a considerable and easily controllable rejection R
an equally large selectivity as that of passive filters known from the state of the art mentioned hereinafter;
a D C. adjustable rejection factor Q1.
The filters according to the invention will be advantageously used in frequency doublers for rejecting the unwanted signal at the fundamental frequency. The circuit according to the invention though being of the second category will not be used as a second order all-pass filter. Actually, it will be noticed that the filter itself is realised by means of all-pass filter elements each of the first order in an embodiment described hereinafter.
On the other hand, also known from the state of the art, there are band-stop filters of the second category operating at high frequencies But these circuits are realised by means of passive elements. For example, the circuit commonly referred to as "Robinson bridge" also known by the name of "Wien bridge", and represented on page 121 of the reference publication entitled "Electric Filters"by T. H. TURNEY (Ed. Pitman and Sons, London, England 1945) could be considered. This circuit, although it is capable of operating at very-high frequencies, has various disadvantages, specifically:
an output that has to be differential,
a high output impedance,
a zero gain, because it is realised by means of passive elements,
generally speaking, difficult control and poor performance
The block diagram of the band-stop filter circuit according to the invention is illustrated in FIG. 1. This circuit is formed by two parts defined by dashed lines. The first part is formed by the building blocks F1 and F2 each having the function of an all-pass filter, and
the second part AS this the function of a summing amplifier.
The blocks F1 and F2 have for their transfer function the resPective relationships (6) and (7) of Table I.
The amplifier block AS is formed by an amplifier A1 having gain k1 and an amplifier A2 having gain k2. The amplifier A1 is arranged in cascade with the filter blocks F1 and F2 while amplifier A2 is arranged in parallel with the branch formed by F1, F2 and A2. The adder S3 sums the output signals of the amplifiers A1 and A2, and furnishes the output signal VS of the circuit according to the invention, whose input signal VE is applied to the junction of the filter block F1 input and the amplifier A2.
Consequently, the transfer function of the whole circuit of FIG. 1 is G(s), provided by the relationship (4), the parameters being defined by the relationships (8), (9), (10), (11) and (16).
In these relationships:
τ1 and τ2 are the respective time constants associated to the poles of the filter blocks F1 and F2 ;
τ'1 and τ'2 are the respective time constants associated to the zeroes of the filter blocks F1 and F2.
It should be observed that in the normal case:
τ'1 =τ1
τ'2 =τ2
and ω1 =ω0.
It should also be observed that the maximum Q0 =1/2 is obtained for
τ1 =τ2.
The circuit according to the invention thus forms a bandstop filter of the second category defined above, which can realise a total rejection.
A first embodiment of the circuit according to the invention is shown in FIG. 2a. This circuit of FIG. 2a is a basic version of the circuit according to the invention, which permits several variants whose characteristic features are shown in the other embodiments.
In the version of FIG. 2a, the all-pass (AP) filter blocks are identical. They each comprise a phase-shifting bridge, which is followed by a differential PUSH-PULL amplifier stage.
Each phase-shifting bridge comprises between the opposite nodes 11, 12 and 21, 22, for filters F1 and F2, respectively, in a first branch a resistor R11, R21, and a capacitor C11, C21 ; and in a second branch a capacitor C12, C22 and a resistor R12, R22. The output signal of each phase-shifting bridge is available at the junctions 13, 14 and 23, 24 for filters F1 and F2 respectively, of the resistors and capacitors of each branch.
The resistors R11, R12, R21, R22 of the phase-shifting bridges are actually, as is represented in the FIGS. 2a, 3a, 3b, the respective field effect transistors (MESFET) T11, T12 for the phase-shifting bridge of the filter block F1, and T21, T22 for the phase-shifting bridge of the filter block F2. The source and the drain of these transistors form the ends of the resistors, and their gate receives a bias voltage E6. These resistors are thus adjustable by controlling this voltage E6 (see FIG. 2a). This control allows setting of the rejection frequency f1 of the circuit according to the invention. A resistor arranged in series with E6 protects the transistors T11, T12, T21, T22 and slightly enhance f1.
The maximum rejection frequency f1 max for the rejection R corresponds to a zero gate-source voltage (VGS =0) for the transistors of the phase-shifting bridges, which is obtained for:
E6 maximum=E2 =2.5V
for the example described above (see FIG. 2a).
The minimum frequency f1 min for the rejection R corresponds to the minimum voltage chosen as a function of the threshold voltage of the transistor that can be applied between the gate and source of the transistors of the phase-shifting bridge, that is to say:
VGS min=-1.8V, which is obtained for E6 min=E2 -1.8V-0.7V in this embodiment.
The ratio of the output voltages VSP of the nodes 13, 14 or 23, 24 to the input voltages VEP of the nodes 11, 12 or 21, 22 of the phase-shifting bridges is provided by the relationship (12) of Table I, in which relationship R0 and C0 are the resistances and capacitances of the different elements of the phase-shifting bridges.
Each filter block F1 or F2, having the function of an all-pass filter, subsequently comprises a differential amplifier stage of which each branch is of the PUSH-PULL type.
These amplifiers, as they are represented in FIG. 2a, comprise in a series arrangement a first branch comprised of a transistor at the top (T13 and T23 for filters F1 and F2, respectively,) L- and a transistor at the bottom (T14 and T24 for filters F1 and F2, respectively). This first branch is arranged in parallel with a second branch comprised of a transistor at the top (T15 and T25, respectively) and in series with a transistor at the bottom (T16 and T26, respectively).
The transistors at the top have common drains which are connected to a D.C. supply voltage E1. The transistors at the bottom have common sources which are connected to ground.
The gates of the transistors at the top are connected to the respective output junctions of the phase-shifting bridge that precedes the amplifier stage, these junctions also being cross-connected to the gates of the transistors at the bottom of the same amplifier stage. Furthermore, the gates of the transistors at the top are biased through a resistor with a D.C. supply voltage E2 (R13 for T13), (R15 for T15), (R23 for T23) and (R25 for T25). The gates of the transistors at the bottom are also biased with a voltage E5 and E4, respectively, for the filters F1 and F2. Isolating capacitors (C13, C14 for F1 and C23, C24 for F2, respectively) are inserted into the cross connections of the transistor gates at the top and at the bottom.
The output signal of each differential amplifier stage occurs between the junctions of the transistors at the top and at the bottom of each PUSH-PULL branch. The output signal of the amplifier stage filter F2 occurs at the junction 25 between the transistors T23 and T24.
On the other hand, the input signal VE of the circuit is applied to the junction 11 of the first phase-shifting bridge through an isolating capacitor C110, the junction 12 receiving a bias voltage E2.
The signals the outputs 13 and 14 of the first phase-shifting bridge are applied to the gates of the transistors 13 and 15 of at the top of the amplifier stage of F1 and to the gates of the transistors T16 and T14 at the bottom of this same amplifier. The signals of the outputs 15 and 16 of this amplifier are applied through isolating capacitors C210 and C220 to the inputs 21 and 22 of the phase-shifting bridge of block F2. The signals of the outputs 23 and 24 of this phase-shifting bridge are applied to the gates of the transistors T23 and T25 at the top and to those of the transistors T24 and T26 at the bottom, cross-connected to the amplifier stage of F2. The signal of the output 25 of this amplifier stage is available at the node of the transistors T23 and T25.
The transistors forming the PUSH-PULL differential amplifier stages are all field effect transistors.
The bias voltage E4 applied to the gates of the transistors at the bottom of the amplifier stage F2 is used for a coarse adjustment of the rejection R of the whole circuit, while the bias voltage E5 applied to the gates of the transistors at the bottom of amplifier stage of F1 is used for a fine adjustment of this rejection R.
The circuit according to the invention furthermore comprises an amPlifier A1 arranged in series with the filter blocks F1 and F2. This amplifier A1, as is represented in FIG. 2a, is formed by the field effect transistor T31.
The source of this transistor T31 is connected to ground and its drain is connected to the supply voltage E1 via a charge. This charge is comprised of the transistor T30 whose drain is connected to the supply voltage E1, whose gate is biased over a resistor R30 by the voltage E2, its gate also being connected to its source across the capacitor C30. The output signal of this amplifier stage A1 is available at the drain of the transistor T31.
The gate of the amplifier transistor T31 (input of the amplifier A1) receives across a capacitor C310 the output signal of the PUSH-PULL amplifier stage of F2 taken off at the node 25 of the transistors T23, T24. On the other hand, this gate is biased by the voltage E1 through a resistor R31.
The bias voltage E4 simultaneously adjusts the gain G2 of the filter block F2 and the gain k1 of the amplifier A1 ; thus, it coarsely adjusts the rejection factor Q1.
Conversely, the bias voltage E5, which is supplied to the gates of the transistors at the bottom of the PUSH-PULL amplifier stage of filter block F1, performs a fine adjustment of the gain of this amplifier stage and thus a fine adjustment of the rejection factor Q1.
In this embodiment we choose:
E1 =6V
E2 =2.5V
E4 will be adjusted for obtaining:
E5 ≃-0.8V.
It should be observed that E2 can be replaced by a resistance bridge inserted between E1 and earth for diminishing the required number of voltage sources.
The circuit according to the invention also comprises an amplifier A2 having gain k2. This amplifier is formed by a field effect transistor T32 arranged in parallel and thus admitting the same charge as the transistor T31 (A1). The amplifier transistor T32 receives on its gate (input A2) the input siqnal VE across an isolating capacitor C320.
The common drains T31 and T32 perform the function of the summation element S3 (see FIG. 1). Besides, the gate of the amplifier transistor T32 (A2) the input signal VE across an isolating capacitor C320.
The common drains T31 and T32 perform the function of the summation element S3 (see FIG. 1). Besides, the gate of the amplifier transistor T32 (A2) is biased by a voltage E3. This bias voltage is to remain fixed, but its value need not necessarily be very accurate (E3 ≃-0.8V) should be used.
The arrangement represented in FIG. 2a adds a phase shift π to the output signal VS available at the common drain of the transistors T31 (A1) and T32 (A2) when compared to the block diagram represented in FIG. 1.
For a bias voltage E6 of the transistor gates of the phase-shifting bridges so that:
E6 =2.5, the rejection frequency is f1 =3.23 GHz,
E6 =0.7V, the rejection frequency is f1 =0.90 GHz.
the frequency band covered is very large.
On the other hand, if the bias voltage E5 is adjusted with an accuracy of ±25 mV, the circuit according to the invention will guarantee:
a rejection R≧56 dB at f1 =3.23 GHz,
a rejection R≧60 dB at f1 =0.90 GHz.
These results are easily obtained, because the circuit according to the invention is particularly easy to adjust. Actually, it is simple to adjust the gain of each PUSH-PULL amplifier stage used. Therefore, in each building block F1 and F2, especially these amplifiers of the PUSH-PULL type have been chosen in preference to the classical inverted amplifiers.
It should also be observed that as the gain of the PUSH-PULL is adjusted better at the lowest frequencies, the rejection there is the best. However, it still remains very high. The low frequency gain of the whole circuit at the lowest rejection frequencies f1 is of the order of 12 dB. It is 10 dB at the highest frequencies f1.
The rejection frequency;
f1 =3.23 GHz is obtained for E4 =-1V and E5 ≃-0.85V
f1 =0.90 GHz is obtained for E4 =-1.6V and E5 ≃-0.95V.
The FIGS. 2b and 2c present by way of example the computerized simulated characteristic values of the ratios of the signals |VS /VE | in dB as a function of the operating frequency, which is expressed in GHz; and the value of the phase shift φ of the signals VS and VE as a function of the frequency expressed in GHz, for E4 =-1V, E5 =-0.85V, E6 -2.5V.
The circuit according to the invention specifically offers the advantage to make f1 adjustable, resulting in:
(a) the possibility to correct the value of f1, of which the integrated circuits always produce dispersed values, owing to the dispersion of the characteristic values of the elements in the various circuits;
(b) the possibility to use the same circuit with many different frequencies.
The Table II presents by way of a non-exhaustive example, values of the elements for the circuit as shown in FIG. 2a, with field effect transistors normally of the band-pass type in the absence of gate-source bias, presenting a threshold voltage of the order of:
VT =-2.5V and a gate length 1=0.7 μm.
The circuit according to the invention is realized by means of elements of Table II on a gallium arsenide substrate (GaAs).
By taking again equation (4) which, as will have been noticed, aPPlies to the present circuit, it has already been observed hereinbefore that: First, the filter circuit according to the invention is of the band-stop filter type if the relationship (13) is satisfied.
The structure according to the invention enables to attain that Q1 is positive or negative and the relationship (14) satisfied. Secondly the circuit according to the invention also enables to attain that Q1 =-Q0, which corresponds to the case in which the circuit has the function of an all-pass filter. In addition, the circuit according to the invention allows to pass from the function of a band-stop filter to the function of an all-pass filter with the same elements.
Actually, it will be sufficient to make the gate bias voltages of the transistors of the different stages vary, to attain the variation
either of k1,
or of k2,
or of G1
or of G2
Then in a continuous manner we can pass from:
Q1 =-∞(band-stop) to Q1 =-Q0 (all-pass).
The second embodiment of the circuit according to the invention is shown in FIG. 3a. The circuit of FIG. 3a is a version that allows a finer adjustment of the gain and thus of the rejection. Consequently, the amplifiers of blocks F1 and F2 are not identical.
The amplifier of filter block F1 does not have the gates of the top and bottom transistors cross-connected, and thus is not of the PUSH-PULL type. It is simply comprised of two followers of which the gain is smaller than in embodiment I, but which can be adjusted more accurately. Thus, the rejection can be adjusted in a simpler manner.
Thus, a choice may be made between the circuit of embodiment I and that of embodiment II according to the intended use.
In this embodiment II a D.C. voltaqe E5 ≃0 is now used, while maintaining therefor negative values (E5 <0). The values of the capacitors of the phase-shifting bridges are different from those of the embodiment 1 as is shown in Table lII.
The rejection frequencies obtained are:
F1 =3.42 GHz, for E6 =2.5V
f1 =0.945 GHz, for E6 =0.7V.
The band covered by the frequency f1 in the circuit of the embodiment II is also a little larger than that of embodiment I.
If the voltage E5 is adjusted to ±25 mV, it is guaranteed that:
R≧59 dB with f2 =3.42 GHz
R≧61 dB with f1 =0.945 GHz.
The gain at the lowest frequencies is equal to 7.5 dB because the followers show less gain than the PUSH-PULL of the previous embodiment, as has already been stated.
To obtain the rejection frequencies
f1 =3.42 GHz, E4 is to be equal to -0.85V, E5 ≃-0.19V
f1 =0.945 GHz, E4 is to be equal to -1.85V, E5 ≃-0.28V.
This embodiment allows differential operation (see FIG. 3b).
The differential inPut signal VE =VE2 -VE1 is applied between the nodes 11 and 12 of the phase-shifting bridge of the first building block F1.
The output signal can be differential: this will be VS =VS1 -VS2, the voltage VS1 being available at node 25 of the differential amplifier of the second filter block F2 and the voltage VS2 at the node 26. The output signal can also be obtained at ground reference; then one of the voltages should be used which is available either at node 25 or at node 26.
The amplifier A1 is combined with the PUSH-PULL amplifier stage of block F2. Thereto, the gates of the bottom transistors T24, T26 of this stage are biased by the same D.C. voltage E4 as the one that biases the gates of the bottom transistors T14, T16 of the PUSH-PULL amplifier stage of filter block F1. This allows coarse adjustment of the rejection of the whole circuit by adjusting E4.
The amplifier A2 is realized by adding field effect transistors T'23, T'24, T'25, T'26 in a parallel arrangement to the respective transistors T23, T24, T25, T26 of the PUSH-PULL amplifier of filter block F2. The gates of the top transistors T'23, T'25 are cross-connected by means of capacitors C'26, C'23 and C'25, C'24 to the gates of the respective bottom transistors T'26, T'24.
On the other hand, these cross-connected gates receive the differential input signal. Then, the gates T'23 and T'26 are connected to the terminal 11 through the D.C. isolating capacitor C110, and the gates of the transistors T'25 and T'24 are connected to the terminal 12 via the D.C. isolating capacitor C120.
The gates of the transistors at the top T'23 and T'25 are also biased by the D.C. voltage E2 through the resistors R'23 and R'25, while the bottom transistors T'24, T'26 are biased by the D C voltage E5 through the resistors R'24, R'26 for the fine adjustment of the rejection.
Table III presents by way of a non-exhaustive example the values of the elements for the use of the circuit of FIG. 3b.
Adjusting the frequency f1 is always effected by having the gates of the transistors of the phase-shifting bridges biased by E6.
The summation S3 is realized because the transistors of amplifiers A1 and A2 are arranged in parallel.
The rejection frequency:
f1 =3.85 GHz is obtained for E6max =E2 =2.5V
f1 =0.906 GHz is obtained for E6max =E2 -1.8V=0.7V.
The frequency band covered by f1 is larger than in the embodiments described hereinbefore. This is because when the frequency f1 is higher, the amplifier stages produce a phase delay leading to a diminishing of the frequency f1. Since in this embodiment III the structure is such that the complete circuit comprises one stage less than in the embodiments I and II, the frequency f1 is less affected by the phase shift caused by the amplifier stages.
If E5, which is the fine adjustment of the rejection, is controlled to ±25 mV, it is guaranteed that:
R≧53 dB with f1 =3.85 GHz
R≧51 dB with f1 =0.906 GHz.
The gain at the lowest frequencies is of the order of 1 dB if the differential output signal VS =VS1 -VS2 is used at the nodes 25 and 26, and approximately -5 dB if an asymmetrical output signal VS1 or VS2 is used at 25 or 26.
On the other hand, the rejection frequency f1 =3.85 GHz is obtained for E4 =-0.8V, E5 ≃-1.09V f1 =0.906 GHz is obtained for E4=-1.95 V, E5 ≃-1.1V with the values of the elements of Table IV.
| TABLE I |
| ______________________________________ |
| ##STR1## (1) |
| p = s/ω1 (2) |
| s = jω (3) |
| ##STR2## (4) |
| ##STR3## (5) |
| ##STR4## (6) |
| ##STR5## (7) |
| ω0 = (τ1 τ2)-1/2 |
| (8) |
| ##STR6## (9) |
| ##STR7## (10) |
| G = k1 + k2 G1 G2 |
| (11) |
| ##STR8## (12) |
| |Q1 | > Q0 |
| (13) |
| 1/Q1 = 0 (14) |
| ##STR9## (16) |
| ______________________________________ |
| TABLE II |
| ______________________________________ |
| (FIG. 2a) |
| Resistors |
| Capacitors Transistors Supply |
| R kQ C11 pF t (λ) |
| μm |
| E V |
| ______________________________________ |
| R13 |
| 10 C11 0.13 T11 |
| 10 E1 |
| 6 |
| R15 |
| 10 C12 0.13 T12 |
| 10 E2 |
| 2.5V |
| R14 |
| 10 C13 1 T13 |
| 40 E3 |
| -0.8 |
| R16 |
| 10 C14 1 T14 |
| 40 E4 |
| R23 |
| 10 C21 0.13 T15 |
| 40 E5 |
| R25 |
| 10 C22 0.13 T16 |
| 40 E6 |
| R24 |
| 10 C23 1 T21 |
| 10 |
| R26 |
| 10 C24 1 T22 |
| 10 |
| R30 |
| 10 C30 1 T23 |
| 40 |
| R31 |
| 10 C310 |
| 1 T24 |
| 40 |
| R32 |
| 10 C320 |
| 1 T25 |
| 40 |
| C110 |
| 1 T26 |
| 40 |
| C210 |
| 1 T30 |
| 25 |
| C220 |
| 1 T31 |
| 35 |
| T32 |
| 10 |
| ______________________________________ |
| E4 = coarse adjustment of rejection R |
| E5 = fine adjustment of rejection R |
| E6 = adjustment of rejection frequency f1 |
| TABLE III |
| ______________________________________ |
| (FIG. 3a) |
| Resistors |
| Capacitors Transistors Supply |
| R kQ C11 pF T (λ) |
| μm |
| E V |
| ______________________________________ |
| R13 |
| 10 C11 0.12 T11 |
| 10 E1 |
| 6 |
| R14 |
| 10 C12 0.12 T12 |
| 10 E2 |
| 2.5V |
| R15 |
| 10 C21 0.12 T13 |
| 40 E3 |
| -0.8 |
| R16 |
| 10 C22 0.12 T14 |
| 25 E4 |
| R23 |
| 10 C23 1.- T15 |
| 40 E5 |
| R24 |
| 10 C24 1.- T16 |
| 25 E6 |
| R25 |
| 10 C30 1.- T21 |
| 10 |
| R26 |
| 10 C31 1.- T22 |
| 10 |
| R30 |
| 10 C32 1.- T23 |
| 40 |
| R31 |
| 10 C110 |
| 1.- T24 |
| 40 |
| R32 |
| 10 C210 |
| 1.- T25 |
| 40 |
| C220 |
| 1.- T26 |
| 40 |
| T30 |
| 45 |
| T31 |
| 10 |
| T32 |
| 70 |
| ______________________________________ |
| E4 = coarse adjustment of rejection R |
| E5 = fine adjustment of rejection R |
| E6 = adjustment of rejection frequency f1 |
| TABLE IV |
| ______________________________________ |
| (FIG. 3b) |
| Resistors Capacitors Transistors Supply |
| R kQ C11 |
| pF T (λ) |
| μm E V |
| ______________________________________ |
| R13 |
| 10 C11 |
| 0.13 T11 |
| 10 E1 |
| 6 |
| R14 |
| 10 C12 |
| 0.13 T12 |
| 10 E2 |
| 2.5 |
| R15 |
| 10 C13 |
| 1 T13 |
| 40 E4 |
| R16 |
| 10 C14 |
| 1 T14 |
| 40 E5 |
| R23 |
| 10 C21 |
| 0.13 T15 |
| 40 E6 |
| R'23 |
| 10 C22 |
| 0.13 T16 |
| 40 |
| R24 |
| 10 C23 |
| 1 T21 |
| 10 |
| R'24 |
| 10 C24 |
| 1 T22 |
| 10 |
| R25 |
| 10 C'23 |
| 1 T23 |
| 30 |
| R'25 |
| 10 C'24 |
| 1 T'23 |
| 10 |
| R26 |
| 10 C'25 |
| 1 T24 |
| 30 |
| R'26 |
| 10 C'26 |
| 1 T'24 |
| 10 |
| C110 |
| 1 T25 |
| 30 |
| C120 |
| 1 T'25 |
| 10 |
| C210 |
| 1 T26 |
| 30 |
| C220 |
| 1 T'26 |
| 10 |
| ______________________________________ |
| E4 = coarse adjustment of rejection R |
| E5 = fine adjustment of rejection R |
| E6 = adjustment of rejection frequency f1 |
| Patent | Priority | Assignee | Title |
| 5124592, | Feb 14 1990 | Kabushiki Kaisha Toshiba | Active filter |
| 5151662, | Jun 24 1991 | Motorola, Inc. | Circuit and method of signal differentiation |
| 5793264, | Oct 06 1995 | Mitel Semiconductor Limited | LAN equalizer |
| 7088833, | Oct 01 1999 | Multiple-speaker | |
| 7656939, | Oct 25 2004 | KAWASAKI MICROELECTRONICS, INC | Adaptive equalizer with passive and active stages |
| Patent | Priority | Assignee | Title |
| 3932819, | Apr 04 1972 | Dolby Laboratories, Inc. | Electrical filters |
| 4137510, | Jan 22 1976 | Victor Company of Japan, Ltd. | Frequency band dividing filter |
| 4388595, | Mar 26 1981 | Bell Telephone Laboratories, Incorporated | Apparatus and method for removing bias distortion from a product demodulator |
| 4716388, | Dec 24 1984 | Multiple output allpass switched capacitor filters |
| Executed on | Assignor | Assignee | Conveyance | Frame | Reel | Doc |
| Oct 24 1988 | U.S. Philips Corp. | (assignment on the face of the patent) | / | |||
| Dec 29 1988 | DEVILLE, YANNICK | U S PHILIPS CORPORATION, A CORP OF DE | ASSIGNMENT OF ASSIGNORS INTEREST | 005013 | /0543 |
| Date | Maintenance Fee Events |
| Oct 18 1994 | REM: Maintenance Fee Reminder Mailed. |
| Mar 12 1995 | EXP: Patent Expired for Failure to Pay Maintenance Fees. |
| Date | Maintenance Schedule |
| Mar 12 1994 | 4 years fee payment window open |
| Sep 12 1994 | 6 months grace period start (w surcharge) |
| Mar 12 1995 | patent expiry (for year 4) |
| Mar 12 1997 | 2 years to revive unintentionally abandoned end. (for year 4) |
| Mar 12 1998 | 8 years fee payment window open |
| Sep 12 1998 | 6 months grace period start (w surcharge) |
| Mar 12 1999 | patent expiry (for year 8) |
| Mar 12 2001 | 2 years to revive unintentionally abandoned end. (for year 8) |
| Mar 12 2002 | 12 years fee payment window open |
| Sep 12 2002 | 6 months grace period start (w surcharge) |
| Mar 12 2003 | patent expiry (for year 12) |
| Mar 12 2005 | 2 years to revive unintentionally abandoned end. (for year 12) |